There were many contradictory evaluation criteria to select next-hop in the delay-disruption tolerance networks(DTN).To solve this problem,an attribute hierarchical model was proposed,in which the predefined criteria were summarized as static identity attributes,forwarding desire attributes and delivery capability attributes(IDC).Based on this model,a novel multi-attributes congestion aware routing(MACAR) scheme with uncertain information for next-hop selection was presented,by adopting an decision theory to aggregate attributes with belief structure and computing partial ordering relations.The simulation results show that MACAR presents higher successful delivery rate,lower average delay and effectively alleviate congestion. There were many contradictory evaluation criteria to select next-hop in the delay-disruption tolerance networks (DTN) .To solve this problem, an attribute hierarchical model was proposed, in which the criteria criteria were summarized as static identity attributes, forwarding desire attributes and delivery capability attributes (IDC). Based on this model, a novel multi-attributes congestion aware routing (MACAR) scheme with uncertain information for next-hop selection was presented, by adopting an decision theory to aggregate attributes with belief structure and computing partial ordering The simulation results show that MACAR offers higher successful delivery rate, lower average delay and effectively alleviate congestion.